Our growth regulation research has been concerned with oncogenes and tumor suppressor genes as positive and negative regulators of normal and neoplastic growth. The work is focused on the regulation of the Ras-superfamily of proteins and their role in cancer, as well as on the role of adhesion in cancer. Here we highlight two ongoing projects. 1) We are studying Sipa-1/Spa-1, which encodes a GTPase-activating protein (GAP) that negatively regulates the Rap-GTPases (a Ras-superfamily group) by enzymatically converting active Rap-GTP to inactive Rap-GDP. In collaboration with the Laboratory of Kent Hunter in CCR, we have identified Sipa-1 as a gene that can promote metastasis in a mouse breast cancer model. Consistent with this role in the mouse model, metatstatic human prostate cancer was found to be associated with Sipa-1 overexpression. Rap-GTP has several functions, including the promotion of cell adhesion and cell polarity, both of which may inhibit metastasis in many tumor systems. Our working hypothesis is that overexpression of Sipa-1 promotes metastasis at least in part by decreasing Rap-GTP-dependent adhesion and polarity. A positive role in cancer has not been reported previously for the negative regulator of a Ras-superfamily group, as most negative regulators of the Ras-superfamily groups inhibit cancer (or have no effect), rather than promote it. If a pathogenetic role is validated for Sipa-1 in human cancer, the gene and its encoded protein may represent an attractive candidate molecular target. In addition, a gene called Anakin has been identified as interacting with Sipa-1, to negatively regulate the Rap-GAP activity of Sipa-1, and to alter the anchorage-independent growth of tumor cells in agar and the tumorigenicity in vivo. A polymorphism of the human Anakin gene has been identified, and found to correlate with disease severity in human breast cancer. 2) We are also investigating E-cadherin, a key cell surface adhesion protein that is a tumor suppressor frequently silenced in many epithelial cancers and melanoma, especially those with high metastatic potential. We have found that E-cadherin specifically inhibits the activation of a variety of receptor tyrosine kinases (RTKs) by their physiologic ligands. In human inflammatory breast cancer, however, E-cadherin is paradoxically elevated. We have found in a human inflammatory breast cancer cell line that the presence of E-cadherin actually sensitizes the cells to ligand-dependent activation of RTKs, which is just the opposite of what is seen in normal cells. The results imply that in inflammatory breast cancer E-cadherin acts paradoxically as an oncogene.